During the past two decades, there has been considerable debate about the professional identity and unique characteristics of osteopathic medicine as a field distinct from allopathic medicine.1–5 Proponents of osteopathic distinctiveness offer three common arguments to support their view: (1) the osteopathic profession's focus on primary care, (2) the use of osteopathic manipulative treatment (OMT)—manually guided forces to improve physiologic function—for various conditions, and (3) claims of a uniquely osteopathic interpersonal manner in physician-patient interactions during medical encounters. Undoubtedly, the osteopathic profession has contributed greatly to the provision of primary care services in the United States. Nationally, 52,827 nonretired osteopathic physicians are practicing, including those in postdoctoral training, and 59% of these practicing osteopathic physicians are in the specialty areas of family or general medicine, internal medicine, or pediatrics.6 Osteopathic physicians provide more than one third of all general and family medicine visits in the Northeast.7
Although historically the quintessential element of osteopathic practice, OMT's role in contemporary osteopathic medicine is decreasing among osteopathic physicians.8 Possible explanations for this phenomenon include (1) a secular time trend and cultural shift away from reliance on manual therapies to full-service health care,3 (2) a more recent trend for osteopathic medical students to spend most of their third and fourth years in hospital systems that deemphasize OMT, (3) growing displacement of osteopathic medical graduates from traditional osteopathic internships and residencies into allopathic graduate medical education programs, and (4) limited time and reimbursement for OMT procedures in a managed care environment. Although research funding is now more readily available from the National Center for Complementary and Alternative Medicine at the National Institutes of Health and from osteopathic philanthropic and professional organizations, OMT's evidence base is still limited despite calls for such research within the profession.9 At present, good evidence for the efficacy of OMT exists only for low-back pain.10
The third claim, the positing of uniquely osteopathic physician-patient interactions, is perhaps the most intriguing and controversial among the claims supporting osteopathic distinctiveness. When surveyed, osteopathic physicians have identified the use of OMT, a caring physician-patient relationship, and a “hands-on” style as their most common differences relative to allopathic physicians.11 Evidence from patient visits supports the claims for such differences when considering the management of musculoskeletal disorders.12 The Maine Osteopathic Outcomes Study found that osteopathic physicians were more likely than allopathic physicians to address the patient's emotional state, discuss preventive measures, and consider family life and social activities.13 The First Osteopathic Survey of Health Care in America (conducted in 1998) found that patients of osteopathic physicians tended to report the greatest levels of satisfaction on such items as wellness, use of educational materials, and time spent with the provider.14 However, data from the National Ambulatory Medical Care Survey (NAMCS) have not shown substantial differences between osteopathic and allopathic physicians with regard to provision of preventive medicine services or time spent with patients.7
The purpose of the present study is to further explore and compare the characteristics of patient visits to osteopathic and allopathic physicians during the provision of ambulatory primary care services at academic health centers (AHCs). Our hypothesis is that physicians (both osteopathic and allopathic) affiliated with AHC sites are more likely to be role models representing the ideals of their respective professions than physicians practicing at non-AHC sites and, consequently, that professional differences should be more apparent at AHC sites.
NAMCS, with its purpose of collecting data on medical care provided in physician offices in the United States, developed more than 30 years ago.15 Detailed documentation of the NAMCS instrument, methodology, and data files that served as the basis for this study is available elsewhere.16–20 NAMCS selects patient visits using a multistage probability sample design to reflect ambulatory medical practice nationwide. The sampling frame for NAMCS includes physicians who met the criteria of being (1) office based, (2) principally engaged in patient-care activities, (3) nonfederally employed, and (4) not in the specialties of anesthesiology, pathology, or radiology.
Patient visits and weights
The basic sampling unit for NAMCS is the office-based physician-patient encounter or “patient visit.” NAMCS assigns each patient visit a weight based on four factors: (1) probability of being selected by the three-stage sampling design, (2) adjustment for nonresponse, (3) adjustment for physician specialty group, and (4) weight smoothing to minimize the impact of a few physician outliers whose final visit weights are large relative to those for the remaining physicians. These patient visits provide unbiased national estimates of ambulatory medical care services and facilitate characterization of such services. National population estimates (NPEs) derived from NAMCS may be unreliable if they are based on fewer than 30 unweighted patient visits or if the relative standard error (SE)—that is, the SE divided by the NPE—is greater than 0.30.16–20
Data collection and processing
Physicians participating in NAMCS or their staff used patient record forms to collect the data for each selected visit. The NAMCS staff performed completeness checks, editing, and quality control measures to ensure the accuracy of patient record forms and associated data files. Item nonresponse rates were 5% or less for most variables. Major exceptions with a higher nonresponse rate included ethnicity (which had a nonresponse rate of 20%), race (18%), tobacco use (30%), and time spent with physician (16%). NAMCS imputed the missing data for some of these—race, ethnicity, and time spent with physician—as well as birth year (4% nonresponse rate) and sex (4%) by assigning the value from a randomly selected patient record form representing another patient with similar known characteristics. NAMCS performed such imputations according to physician specialty, geographic region (using state instead of geographic region to impute ethnicity), and primary diagnosis codes.
Data management and statistical analyses
We acquired the electronic files containing the 2002 through 2006 NAMCS data from the National Center for Health Statistics. We merged and analyzed the files using SPSS Version 15.0 for Windows (SPSS Inc., Chicago, Illinois). Because the multistage probability design of NAMCS includes clustering, stratification, and the assignment of unequal probabilities of selection to sample units, we performed all analyses with the SPSS complex samples module to accurately compute NPEs and their SEs.21 The Office for the Protection of Human Subjects at the University of North Texas Health Science Center certified that this research was approved and exempt from full board review by the institutional review board.
This study focused on primary care patient visits to physicians at AHCs compared with those at non-AHC sites. We identified AHCs using two survey items: (1) “Who owns this practice?” and (2) “Type of office setting for this visit.” We designated AHC visits as those in which physicians indicated either that a “medical/academic health center” owned their practice or that a “faculty practice plan” was the setting for the patient visit. Primary care specialties included general and family medicine, internal medicine, and pediatrics.
Patient sociodemographic characteristics included age, sex, race, and ethnicity. Patient visit context characteristics included geographic region, metropolitan statistical area (MSA) status of the practice, whether symptoms were acute or chronic, and whether the visit was related to an injury. We proxied the last using the survey item which asked whether the visit was related to an injury, poisoning, or adverse effect of medical treatment (IPA), and we corroborated that 88% of the IPAs specifically described on another open-ended item in the NAMCS database were injuries rather than poisonings or adverse effects. Physician provider characteristics included physician specialty, primary care physician status, and type of physician provider (doctor of osteopathy [DO] or doctor of medicine [MD]). The elements of medical management included any diagnostic tests, patient counseling, or drugs that the provider ordered. In this study, “ordered” meant that the provider ordered, scheduled, or performed the element of medical management during the patient visit. NAMCS broadly defines drugs as any medications or injections, including immunizations, allergy shots, anesthetics, or dietary supplements, that a provider ordered, supplied, administered, or continued during the visit, regardless of prescription or over-the-counter status.
To study clinical practice characteristics at AHC sites, we initially derived and stratified overall NPEs of patient visits according to patient sociodemographic characteristics, patient visit context, physician provider characteristics, and medical management elements within several subgroups (Figure 1). We then used multiple logistic regression to compute adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for factors associated with patient visits to osteopathic physicians at non-AHC sites (versus allopathic physicians at non-AHC sites) and with patient visits to osteopathic physicians at AHC sites (versus allopathic physicians at AHC sites). We used general linear models to compare these subgroups of patient visits with regard to the amount of time spent with the physician and the number of drugs ordered. Finally, we used contingency table methods and adjusted ORs and 95% CIs to estimate OMT use by osteopathic physicians. In NAMCS, physiotherapy includes treatments using heat, light, sound, or physical pressure or movement (e.g., “manipulative therapy”) during a patient visit, and thus it served as a surrogate for estimating OMT use. Assuming comparable frequencies of non-OMT physiotherapies by osteopathic and allopathic physicians, increasing magnitudes of the OR (>1) are indicative of greater OMT use by osteopathic physicians. Although not specific for OMT, this analytical approach has been successfully used with NAMCS data to demonstrate the distinctiveness of osteopathic physicians in treating musculoskeletal disorders.12 We tested all hypotheses at the .05 level of statistical significance.
Survey responses and national population estimates of patient visits
There were 134,369 patient visits (NPE, 4.57 billion patient visits; SE, 220.2 million patient visits) represented in the entire NAMCS database for 2002–2006. The flow chart illustrating how we derived the subgroups that formed the basis for our study is shown in Figure 1. We designated a total of 6,358 patient visits (NPE, 107.3 million patient visits; SE, 14.0 million patient visits) as having occurred at AHC sites. These included 3,825 patient visits (NPE, 58.1 million patient visits; SE, 8.2 million patient visits) involving primary care physicians in the specialties of family medicine, internal medicine, or pediatrics. A total of 107.3 million of all 4.57 billion (2.3%) patient visits were provided by physicians (DO or MD) at AHC sites, including 2.9% of all primary care visits. Osteopathic physicians provided 10.0% of primary care patient visits at AHC sites, compared with 10.4% of primary care patient visits at non-AHC sites. The overall characteristics of the 40,701 primary care patient visits (NPE, 2.0 billion patient visits; SE, 100.0 million patient visits) are presented in Table 1.
Comparison of primary care patient visits provided by osteopathic and allopathic physicians at non-AHC sites
As a frame of reference, Table 2 presents the differences between osteopathic and allopathic physicians in the provision of primary care at non-AHC sites. Compared with allopathic physicians, osteopathic physicians were less likely to provide primary care for the following demographic groups: patients in all age categories other than 25 to 44 years of age, blacks and other racial minorities, and Hispanics. Within the context of the patient visit, osteopathic physicians were more likely than allopathic physicians to see patients who reported an injury as the reason for their visit. During the provision of primary care, osteopathic physicians at non-AHC sites ordered more diagnostic tests than allopathic physicians.
Comparison of primary care patient visits provided by osteopathic and allopathic physicians at AHC sites
In contrast to the differences between osteopathic and allopathic physicians in non-AHC sites described above, Table 2 shows other important differences between osteopathic and allopathic physicians at AHC sites. Osteopathic physicians at AHC sites were less likely to provide primary care for females and were more likely to provide such care for patients from non-MSAs than allopathic physicians at AHC sites. At AHC sites, osteopathic physicians were less likely to provide chronic medical care and to perform diagnostic tests than allopathic physicians. However, osteopathic physicians were more likely to order drugs than their allopathic counterparts at AHC sites. The results also showed a trend toward decreased likelihood of osteopathic physicians providing primary care patient visits for blacks and other racial minorities at AHC sites.
Time spent with physician during patient visits
Overall, patients spent a mean 17.49 minutes (95% CI, 17.05–17.92 minutes) with their physician (DO or MD) during primary care patient visits in the specialty areas of family medicine, internal medicine, or pediatrics. Osteopathic physicians at non-AHC sites spent less time with patients than allopathic physicians at non-AHC sites (P = .01). The provider- and site-specific mean times spent with physician were
* 16.41 minutes (95% CI, 15.64–17.18 minutes) for osteopathic physicians at non-AHC sites;
* 17.61 minutes (95% CI, 17.13–18.09 minutes) for allopathic physicians at non-AHC sites;
* 15.44 minutes (95% CI, 12.11–18.77 minutes) for osteopathic physicians at AHC sites; and
* 17.70 minutes (95% CI, 16.23–19.17 minutes) for allopathic physicians at AHC sites.
Number of drugs ordered during patient visits
Overall, the mean number of drugs physicians (DO or MD) ordered was 2.20 (95% CI, 2.11–2.29) during primary care patient visits in the specialty areas of family medicine, internal medicine, or pediatrics. Osteopathic physicians at AHC sites ordered more drugs than osteopathic physicians at non-AHC sites (P = .001) and than allopathic physicians at AHC sites (P = .02). The provider- and site-specific mean numbers of drugs ordered were
* 2.34 (95% CI, 2.17–2.52) for osteopathic physicians at non-AHC sites;
* 2.17 (95% CI, 2.08–2.27) for allopathic physicians at non-AHC sites;
* 3.24 (95% CI, 2.72–3.75) for osteopathic physicians at AHC sites; and
* 2.46 (95% CI, 1.99–2.92) for allopathic physicians at AHC sites.
Osteopathic manipulative treatment during patient visits
At non-AHC sites, osteopathic physicians reported using physiotherapy during 166 (3.0%) patient visits (NPE, 6.1 million patient visits; SE, 1.6 million patient visits), whereas allopathic physicians reported used physiotherapy during 360 (1.2%) patient visits (NPE, 21.0 million patient visits; SE, 2.1 million patient visits) (P = .03). Thus, the greater reported use of physiotherapy by osteopathic physicians at non-AHC sites can be attributed to OMT in at least 1.8% of patient visits. The corresponding adjusted OR was 1.87 (95% CI, 1.09–3.22). At AHC sites, osteopathic physicians reported using physiotherapy during only 1.2% of patient visits compared with 1.5% of patient visits reported by allopathic physicians. However, the reported levels of physiotherapy use at AHC sites were not adequate for reliable statistical comparisons.
This study is unique in providing a nationally representative perspective on the clinical practice characteristics of patient visits at AHC and non-AHC sites. Overall, about 3% of primary care patient visits were provided by physicians at AHC sites, representing almost 12 million patient visits annually in the United States. An osteopathic physician provided the care for approximately 1 of every 10 of these patient visits. A common finding for both osteopathic and allopathic physicians at AHC sites was that they provided a greater percentage of patient visits in non-MSA settings than their counterparts in non-AHC sites. These findings may reflect movement among academic units, as a managed care strategy, toward the use of more geographically dispersed community-based physicians.22
The study findings for primary care patient visits generally did not support our hypothesis that osteopathic physicians affiliated with AHC sites would serve as role models by exhibiting clinical practice patterns characteristic of a distinctive osteopathic approach to medical care. Osteopathic physicians at AHC sites spent less time with their patients and ordered drugs more frequently than their allopathic counterparts. Osteopathic physicians at AHC sites did not provide patient counseling more frequently than allopathic physicians at AHC sites. The last is consistent with the previous NAMCS finding that osteopathic physicians in general and in family medicine did not provide more frequent patient counseling relating to nutrition or diet, weight reduction, exercise, tobacco use, or mental health or stress reduction than their allopathic counterparts.7
When compared with allopathic physicians at AHC sites, osteopathic physicians at AHC sites spent less time with patients (more than two minutes less on average). One possible explanation could be that more medical students and residents are present during patient visits at allopathic AHC sites compared with osteopathic AHC sites, thereby creating greater opportunities for patient counseling and increasing the time spent with physicians at allopathic AHC sites. Further research is needed to determine whether critical structural characteristics or processes of care impact the time osteopathic physicians spend with patients at AHC sites.
Several of the differences in clinical practice characteristics between osteopathic and allopathic physicians that we observed in non-AHC sites, such as differences in patient demographic characteristics, were merely reflective of the greater statistical power associated with the number of patient visits at non-AHC sites compared with AHC sites. However, we did detect two fundamental differences between osteopathic and allopathic physicians in non-AHC sites that we did not observe at AHC sites. First, at non-AHC sites, osteopathic physicians were more likely than allopathic physicians to see patients who reported an injury as the reason for their visit. This represents a reversal of the findings regarding patient visits at AHC sites; the mild to moderate musculoskeletal injuries typically managed by community-based osteopathic physicians may explain the difference. Second, and fundamentally more important, osteopathic physicians at non-AHC sites were more likely than allopathic physicians to report using physiotherapy, an OMT surrogate measure. Our results show no corresponding evidence of increased physiotherapy use among osteopathic physicians at AHC sites relative to allopathic physicians at AHC sites. One possible explanation we have considered for the presumed underutilization of OMT by primary care osteopathic physicians at AHC sites may be that they referred patients who needed OMT to readily available and better trained specialty physicians in osteopathic manipulative medicine at osteopathic AHC sites. Because we did not consider osteopathic manipulative medicine specialists to be primary care physicians in our study, we would not have captured the use of OMT by such specialists. Osteopathic physicians at non-AHC sites may have had more limited access to osteopathic manipulative medicine specialists for patient referrals, and thus they may have more often provided OMT themselves. Less OMT use by osteopathic physicians at AHC sites may also help explain why they spent less time with patients than osteopathic physicians at non-AHC sites.
We should note several limitations of this study. We made simplifying assumptions in certain analyses because of constraints inherent in the NAMCS patient record form. We based the AHC status of physicians on self-reporting regarding ownership of their practices and the type of office setting for patient visits. To independently confirm AHC status in our analyses was impossible. Also, some pairings of physicians and AHC sites may have been discordant. For example, osteopathic physicians may have been affiliated with allopathic AHC sites and vice versa. Because we cannot measure the existence and degree of such crossover with the available NAMCS data, we could not assess its impact on the study results.
We assessed each of the elements of medical management (i.e., diagnostic tests, patient counseling, and drugs) with one survey item that asked whether the provider “ordered, scheduled, or performed” the element. However, we were unable to verify whether the reporting physician or somebody else performed the reported elements within the relevant patient visit or at any other time. Similarly, we contended with a lack of specificity in measuring the use of OMT because the surrogate physiotherapy variable also included other treatments involving heat, light, sound, or physical pressure or movement. Nevertheless, our conclusion that the increased reporting of physiotherapy by osteopathic physicians at non-AHC sites reflects increased use of OMT rather than physical therapy is consistent with the finding that osteopathic physicians use OMT as a substitute for other therapies, including drugs and physical therapy.23 The “number of drugs ordered” survey item included both prescription and over-the-counter drugs (including vitamins and dietary supplements). Thus, the ordering of a relatively large number of drugs during a patient visit may reflect clinical practice patterns ranging between two extremes: health promotion through the use of over-the-counter vitamins and dietary supplements, and overreliance on prescription drug use. Patient severity of illness may have potentially confounded the results of our study with regard to differences between osteopathic and allopathic physicians. Although the NAMCS patient record form provided reasons for the patient visit and diagnoses, it did not provide a global assessment of severity of illness or general health status to use for adjustment purposes.
Results based on fewer than 30 NAMCS patient visits or with SE greater than 30% of the NPE should be interpreted with caution because of lack of precision. We were generally able to overcome this limitation by combining NAMCS data for five years to achieve a larger sample size for analysis. For some variables (e.g., race), we combined certain survey response items (e.g., “other”) to alleviate this problem; however, such aggregation limits our ability to draw conclusions about the combined categories (e.g., Asian, native Hawaiian/other Pacific Islander, and American Indian/Alaska Native).
In summary, as reflected by spending more time with patients and using OMT more frequently, evidence suggests a more distinctive osteopathic approach to primary care in the community, non-AHC setting than at AHC sites. These findings have potentially important implications for osteopathic medical education; we cannot overstate the impact of role modeling and mentoring by faculty at osteopathic medical colleges and in postgraduate training programs. As ever-increasing numbers of osteopathic physicians continue to enter allopathic graduate medical education programs24 and, eventually, the ranks of faculty at the colleges of osteopathic medicine, distinctive aspects of osteopathic medical practice are likely to diminish in the absence of a concerted professional response. Additional research is needed to better determine the structural characteristics and processes of primary care, including physician-patient interactions, at osteopathic AHC sites. If OMT is to remain a distinctive aspect of osteopathic practice, a greater integration of OMT within primary care at osteopathic AHC sites seems necessary.
Grants from the Osteopathic Heritage Foundation and from the National Center for Complementary and Alternative Medicine of the National Institutes of Health (grant number K24-AT002422) supported Dr. Licciardone's work on this project.
Neither the Osteopathic Heritage Foundation nor the National Center for Complementary and Alternative Medicine participated in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or writing, review, or approval of the manuscript.